1

Fuzzy Logic and Fuzzy Algorithms CISC871/491

Md Anwarul Azim

(10036952)

2

Presentation Outline

� Fuzzy control system

� Fuzzy Traffic controller

� Modeling and Simulation

� Hardware Design

� Conclusion

2

3

Fuzzy Control System

4

Fuzzy Control

Fuzzy control provides a formal methodology for representing, manipulating, and implementing a human’s heuristic knowledge about how to control a system.

Figure from Prof. Emil M. Petriu, University of Ottawa

3

5

Advantage

Advantages of FLC:• Parallel or distributed control multiple fuzzy rules – complex nonlinear system• Linguistic control. Linguistic terms - human knowledge• Robust control. More than 1 control rules – a error of a rule is not fatal.

Useful cases:

� The control processes are too complex to analyze by conventional quantitative techniques.

� The available sources of information are interpreted qualitatively, inexactly, or uncertainly.

� Embedded design & control, Industrial control, Robot control etc

6

Basic Structure of Controller

DEFUZZIFIER – It extracts a crisp

value from a fuzzy set.

· Smallest of Maximum.

· Largest of Maximum.

· Centroid of area.

· Bisector of Area

· Mean of maximum.

FUZZIFIER

Fuzzifier takes the crisp inputs to a fuzzy controller

and converts them into fuzzy inputs.

FUZZY RULE BASE (Knowledge base)

It consists of fuzzy IF-THEN rules that form the

heart of a fuzzy inference system. A fuzzy rule

base is comprised of canonical fuzzy IF-THEN

rules of the form IF x1 is A1(l) and ... and xn is An

(l)THEN y is B(l), where l = 1, 2, ...,M. Should have

Completeness, Consistency, Continuity..

FUZZY INFERENCE ENGINE

Fuzzy Inference Engine makes use of fuzzy logic

principles to combine the fuzzy IF-THEN rules.

Composition based inference (Max/Min,

Max/Product) and individual-rule based inference

(Mamdani). Other methods like Tsukamoto,

Takagi Sugeno Kang (TSK)

4

7

Example

“Fuzzy Control” Kevin M. Passino and Stephen Yurkovichhttp://if.kaist.ac.kr/lecture/cs670/textbook/

8

Fuzzy Traffic controller

A Design Methodology for the Implementation of a Fuzzy LogicTraffic Controller Using FPGA Technology

Mandar Ambre, Bing W.Kwan and Leonard J.Tung

Department of Electrical & Computer Engineering

FAMU-FSU College of Engineering, Florida State UniversityTHE HUNTSVILLE SIMULATION CONFERENCE HSC 2003

5

9

Overview

Why Fuzzy traffic controller--No obvious optimal solution--Most traffic has fixed cycle controllers that need manual changes to

adapt specific situation--One of the desirable features of traffic controllers is to dynamically

effect the change of signal phase durations--This problem can be solved by use of fuzzy traffic controllers which

are capable of signaling adaptively at an intersection.

Goal--Improving Safety--Balanced traffic flow--Minimizing travel time--Increasing the capacity

10

Traffic ControllerCase example

6

11

Membership functions and Fuzzy Rule Base

12

MATLAB

http://www.mathworks.com/help/toolbox/fuzzy/fp243dup9.html

7

13

Modeling example (Simulink)

http://www.youtube.com/watch?v=hFWGToL-NHwhttp://www.mathworks.com/products/simulink/demos.html

14http://www.youtube.com/watch?v=hFWGToL-NHwhttp://www.mathworks.com/products/simulink/demos.html

Modeling using Simulink (Cont.)

8

15

Hardware Design

16

Conclusion

�Overview of Fuzzy Controller basic theory

�Case Study of Fuzzy Traffic Controller Desin

�Discuss about Simulation Environment

�Hardware Implementation Issue

Further Improvement

•Evolving Fuzzy system

•Utilize genetic algorithm

•Use Dynamic inputs

•Advanced EDE tools (H/W)

Thank you!

9

17

Case Study 2 (Extra)